jie zeng, matahel ansar, zubayed rakib, seyed hajimirzaie

Jie Zeng, Matahel Ansar, Zubayed Rakib, Seyed HajimirzaieApplied Hydraulics, Hydrology and Hydraulics Bureau

South Florida Water Management District

Background Great deal of hydraulic designs are carried out in support of the Everglades

Restoration Projects. Reduced-scale physical models typically implemented: reliable but costly.

Computational Fluid Dynamics (CFD): Evaluate and optimize hydraulic performance and design of hydraulic structures in Everglades Restoration projects

Governing equations, NS :

Turbulence model: k-ε closure

qxu

j

j =∂∂

)]

([

) (

) (i

ij

i

i

je

jji

j

i Fx

gxu

xu

xuu

xtu

+−+=+∂∂ζρ

∂∂

∂∂

µ∂∂ρ

∂∂

∂ρ∂

) (

) (

)( ρε

∂∂

σµ

∂∂ρ

∂∂

∂ρ∂

−+=+ kjk

e

jj

j

Gxk

xku

xtk

)() (

) (

)(

21 ρεε∂ε∂

σµ

∂∂ερ

∂∂

∂ρε∂

ε

CGCkxx

uxt k

j

e

jj

j

−+=+

Commercial CFD-software package ANSYS FLUENT

Case Study I: S333N Spillway Design, Layout and Impact Assessment

S333 is a trapezoidal-sill reinforced concrete spillway, located at the intersection of L-29 and L67 canals

Proposed new S333N spillway to accommodate additional discharge

Objective:Determine the layout, the design, operation criteria, and impact of a newly proposed spillway


Layout Alternatives

South Alternative L-67

L-29

North Alternative L-67

L-29

S333 Capacity: 1,350 cfs, One gate 29 ft wide

S333N Proposed Capacity: 1,150 cfs, Two gates each 14 ft wideS333N Required gate opening: 2 x 6.40 ft at design HW of 9.5 ft-NGVD, and TW of 9.0 ft-NGVD)

S333N Sizing

3/13/2

3/2

gLQyc =

b

c

GhHa

Gy

−=

00

0.10

≥Gh


Flow Scenario A: 75% flow from L-67

Near Bed Velocity Contours

South alt: 1.0-2.0 ft/s North alt: 1.0-3.0 ft/s

Limestone layer: scouring not likely

Flow Scenario B: 50% flow from L-67

South alt: 1.5-3.2 ft/s North alt: 1.8-3.2 ft/s

Eddy formation downstream, Flow bias towards east bank in L-29 CanalPlace proposed S333N structure north of the existing

structure S333 at angle 25-30 degrees with S333

H=9.5 ft, T=9.0 ft


Further Analysis Extreme scenarios analysis: high flows + low tailwater

With the adjusted angle of S333N spillway, flow jets are evenly distributed at downstream, without any severe potentials of eddy formation or scouring

As conditions became extreme, flow jet downstream of the structures began to oscillates between north and south bank of L-29 Canal

H=10.5 ft, T=8.5 ft, Q333=1,350 cfs, Q333N=1,150 cfs


Near Bed Velocity

Design Optimization

Split Island

S333N

S333

Split Island

Sheet Pile

S333N

S333

Flow

Flow Deflector

Add flow Deflector

S333N

S333

S333N

S333o Add split Islando Add flow deflectoro With/without sheet pile

Split Island

S333N

S333

Near Bed Flow Field

Near Bed Velocity (ft/s)

QS333N 1150 cfs, QS333 1350 cfsHW 10.5 ft, TW 8.5 ft-NGVD

Split island (1V:2H)Sheet pile

Conceptualsplit island

Split island slope (1V:2H)

Max. Near bed velocity 3 ft/s

Surface Flow Field

Split Island

S333N

S333

Surface Level Velocity (ft/s)

QS333N 1150 cfs, QS333 1350 cfsHW 10.5 ft, TW 8.5 ft-NGVD

Split island (1V:2H)Sheet pile

ConceptualSplit island

Split island Slope (1V:2H)


Design Improvements Installation of flow deflectors at both end-sills, raised by 1.5-2 ft

Flow jet travels longer and expands slower for energydissipation without the flow deflector

Deflector directs the discharge upwards, reduces near bed velocities

Near bed velocities significantlyreduce from 4 ft/s to 2.5 ft/s

Reduces riprap protectionrequirements in L-29 canal


Conceptual Flow Deflector

Without Deflector

With Deflector

Case Study II: S332B and C Pump Stations Refurbishment Design

S332B/C pump stations are located south of Pump Station 331, along the L-31N canal

Construction did not adhere to District standards, meant to be temporary, Frequent repair works

Inflow canal leading to the pump is oriented at 90o with the pump: flow field biased

Objective:Apply CFD model to optimize the refurbishment of S332B/C Pump Stations for improving hydrodynamic performance

S332B Layout Alternatives

1) Move pump Downstream2) Add vanes3) Move pump further west

Alternative 2


1.5 ft NGVD

10 ft NGVD

3.0 ft NGVD

Hist @ 3.69 ft NGVD

ERTP @ 5.0CEPP @ 5.0SD I @4.5

CEPP @ 8.39ERTP @ 8.30SD I @8.22 Hist@ 9.88

SDI @ 7.80CEPP @ [email protected] Hist@ 9.45

S332BFlow simulation uses H=3.00 ft, T=10.0 ft NGVD


With VanesMesh

• Canal bottom @ -10 ft-NGVD based on as-builts

• Forebay extended 50 ft• Slope 1:10 near forebay• 4 Diesel pumps (125 cfs)• 2 Electric pumps (75 cfs)• Design Capacity 650 cfs, the

bottom elevation is -12.5 ft NGVD

Diesel pumps

Electric pumpsSlope 1:10

Without Vanes

-10 ft-NGVD

3 ft-NGVD

-12.5 ft-NGVD

1.5 ft-NGVD


S332B

ExistingCondition

V_mag (ft/s)

Flow biased at bend

Flow biased at intake


Flow vane improves pump approach flow distributions

Proposed Condition Simulation with and Without Vanes


Proposed Condition: Simulation with Vanes and Trash Rack

Summary

CFD successfully applied to hydraulic analysis of two water control structures in Everglades Restoration Projects

CFD is used as a complement or alternative to physical model and prototype results

CFD was systematically used to: o Evaluate structure performance and designo Predict flow behavior and operation risko Optimize structure design

jie zeng, matahel ansar, zubayed rakib, seyed hajimirzaie

Documents